Re: Radon Exposure Assessment for Cases

[Kai Kaletsch <info@eic.nu>]

Sorry Don,



My comment about causality was not very clear:



There are 2 ways of increasing the radon levels in a house: 1) increase the

source term and 2) reduce the ventilation. An increased source term would

most likely show up as high basement readings while reduced ventilation

would show up as a smaller difference between basement and main floor.



If the cause for excess cases was high cumulative radon exposures (let's

assume radon exposure is 100% determined by the main floor concentration), I

would expect some of the high exposures (high main floor readings) to be

caused by high source terms and some of the high exposures to be caused by

low ventilation. That means that I would expect both higher average basement

readings AND smaller differences between basements and main floors.



Looking at the Iowa data, it looks like only the second mechanism (low

ventilation) was causing the higher main floor radon levels. This has me a

bit stumped, because I would think that there would be some differences in

source terms across Iowa. These differences do not show up in the data.



If you take the data at face value, you would have to conclude that some

other factor related to ventilation was the cause for excess cases, not

radon.



There is only a 12% difference in the main floor radon levels between cases

and controls, but there is a 69% difference in the difference between

basement and main floor levels between cases and controls. If I were to

hypothesize on a causal mechanism, I'd pick the 69% effect (ventilation) not

the 12% effect (radon).



> How would I explain it?  It looks like the controls had slightly larger

> houses (more cubic feet) so there could possibly be more dilution of the

> radon in the upper floors.



It would have to be a fairly large difference in house size to explain the

values. We have a model for that situation on our web site (

http://members.shaw.ca/eic/Tools/TwoRoomRadonShort.htm ). You can fool

around with it and see the size difference that is required to explain the

data. (Let the basement be the first chamber and the main floor be the

second chamber.) Was there no matching done for socio-economics?



> Who knows, I would not put to much emphasis on

> averaged data.



I think when you assume LNT, averaged data is quite important because

average risk is supposed to be proportional to average exposure. If the

categorical analysis disagrees with the analysis on averages, it is likely

due to the choices in picking the categories. (It doesn't matter if these

were chosen a priory or not.)



Kai



----- Original Message -----

From: "Rad health" <healthrad@hotmail.com>

To: <info@eic.nu>; <radsafe@list.vanderbilt.edu>

Sent: Saturday, January 26, 2002 12:27 PM

Subject: Re: Radon Exposure Assessment for Cases





>

>

>

> >From: Kai Kaletsch <info@eic.nu>

> >When too many smokers come to visit, I will turn on the bathroom fan to

> >increase ventilation.

>

>

> Kai, interesting comments.

> I doubt people would come to visit people with lung cancer and then smoke,

> so they would have less ventilation than controls, right?

>

>

> >Assuming that someone with lung cancer has the same lifestyle as someone

> >without, seems to be a pretty big "what if".

>

> I think Iowa did some second and third year radon measurements when the

> subjects became more ill, they reported the year to year variation in the

> paper.  They could probably look at that.  But, remember, there are other

> people living in the house so the lifestyle factors may not be all that

> different.

>

> This is likely more of a problem in studies like Martin's where they go

back

> to houses, to measure radon, that have not been occupied by the subject

for

> many years.  It would be interesting to hear how many of the homes in

> Martin's study are occupied by the subject when they performed the radon

> measurements.  In the Iowa study, the data looked like there was more of a

> change in radon where houses were now occupied just by next of kin.  Can

you

> imagine the changes that occur in homeowner behaviours from one family to

> the next?  I think this is why the Iowa Study tried to maximize the number

> of living subjects.

> >

> >Don, how do you explain the fact that cases and controls had similar

radon

> >levels in the basement (actually controls had slightly higher levels),

but

> >cases had higher levels in the living areas?

>

> I think your talking about the geometric mean levels? I don't know how

much

> averaged data will tell you. I am more interested in the cumulative

> exposures weighted by how much time was spent in the house.  Didn't they

do

> a risk plot if just basement measurements were used?  I recall it still

> showing an upward direction.  No one spends time in the basement so that

> information is not that informative about risk.

>

> How would I explain it?  It looks like the controls had slightly larger

> houses (more cubic feet) so there could possibly be more dilution of the

> radon in the upper floors.  Who knows, I would not put to much emphasis on

> averaged data.

>

> >

> >If cumulative radon exposure was really the causal factor, I would expect

> >at

> >least some excess cases to be caused by a high radon source term (high

> >basement reading). What the data seems to suggest, however, is that all

the

> >excess cases are due to poor ventilation (small difference between

basement

> >and living areas).

>

>

> The subjects spent very little time in the basements so that is really not

a

> significant factor.  You are also may be ignoring the fact the WLMs were

> linked to where the subject spent time within the home and how much time

was

> spent in the home.  After all, exposure = time at a certain concentration.

>

>

> I understand the challenge of trying to decide between Cohen's data and

the

> Iowa Data.  I think the suggestion Dr. Field made to Dr. Cohen would help

to

> improve the quality of his analysis.

>

> Don

>

> _________________________________________________________________

> Join the world's largest e-mail service with MSN Hotmail.

> http://www.hotmail.com

>



************************************************************************

You are currently subscribed to the Radsafe mailing list. To unsubscribe,

send an e-mail to Majordomo@list.vanderbilt.edu  Put the text "unsubscribe

radsafe" (no quote marks) in the body of the e-mail, with no subject line. You can view the Radsafe archives at http://www.vanderbilt.edu/radsafe/